Wind pollination, clonality, and the evolutionary maintenance of spatial segregation of the sexes

Abstract

Spatial segregation of the sexes (SSS) has been reported in >30 dioecious species from 24 plant families. Such convergent evolution can arise via shared constraints and/or shared selective pressures but the extent to which SSS is adaptive is unresolved. In particular, because mating in plants occurs primarily among near neighbors, any fitness benefits of SSS should usually be offset by the separation of mating partners during reproduction. It has been argued that the problem of mating in plants with SSS should be alleviated in plants with more reliable modes of pollination, such as wind pollination. Accordingly, we evaluated the association between SSS and pollination mode using phylogenetically controlled comparative analyses. Moreover, because outcrossing plants produce many more pollen grains than ovules, females should compete less intensely for mating partners than males. We therefore predicted that any fitness costs arising from segregation should be greater for males than females. We evaluated this prediction using spatially explicit stochastic simulations. The comparative analyses indicated that there is indeed a positive association between wind pollination and SSS, but also that SSS might be associated with clonal growth. Because methods used for detecting SSS can be biased by clonality, this raises the possibility that some cases of SSS are ‘false positives’. The simulation results supported the prediction that male fitness is more strongly affected by segregation than female fitness. Contrary to arguments based on anecdotal evidence, increasing the effectiveness of pollen dispersal had little effect on the fitness of segregated males. Instead, our results indicated that the evolutionary maintenance of SSS requires substantial performance advantages to segregated males. However, there is little empirical evidence that such performance advantages exist. Future efforts to understand the evolution of SSS should focus on the detection of performance differences between segregated and unsegregated plants, particularly males.